not worth it unless you can stick another radeon 57xx or better in there.

There's 26 characters in the alphabet, so that gives us 52 characters upper + lower case. Then add in 0-9 which totals 62 characters for any place in the address. The address is 34 characters long. Thus, we have 62 possibilities on 34 places, or 62^34 different possibilities of addresses (8.7*10^60). In other words, they aren't going to run out, and the chances of having two identical addresses are practically 0, even if there are billions of billions (10^18) of addresses in the wild which will take practically infinity to create (even at 1 billion addresses per second, it would take 30 years).

The workaround is the currently accepted method of making sure that the sender is the right person, and from a technical standpoint it will work for as long as BTC is around.

Eligius is a little hit or miss - its hashrate is too low to give a consistent payout.

I've found that slush's pool has given me the most consistent and best payout...might be luck, but that's the one I'm sticking with.

They're both grounded at the socket, so no. It's good practice to have them share a ground connection though.

It's pretty easy hack them together with just a little bit of wire. All you have to do is draw 1 wire (pin 14, green wire) from one ATX connector to the other (where the "on/off" signal is generated). The PSU knows to turn on when this wire's voltage goes to 0 (ground). The adapters also include a ground wire (black) which isn't strictly necessary, but does insure both PSUs are grounded correctly (safety thing, not strictly necessary). You can draw this wire as well since it's about a 2 minute job.

This is my point exactly - people who are not efficient will be squeezed out of the market pretty quickly.

Also, as I laid out in my post, the only reason we would see an uptrend in BTC is that demand continues to grow. The currency's value is entirely demand driven.


I gave the 2x5870 as an example of a rig that would get around 2mhash/w when properly configured. It's the most power efficient card for mining.

You happen to live in an area with cheap electricity, so mining will be more profitable for you in the long run. The national average for electric is around $.15/kwh which is why I chose that figure.

If things keep going as they have been for the last half year, we're looking at 7 more two week cycles until mining becomes completely unprofitable for pretty much everyone but the most efficient people. This happens roughly around difficulty 10 million given that BTC price stays at $20 or less.

Given that we are now at difficulty .877 million, and the average rate of difficulty increase over the last 8 months has been about 55% per two weeks, we will reach difficulty 10mil+ in 13 weeks. At that point an electricity cost of just $.15 / KWH is going to be enough to make mining unprofitable on a machine that nets ~2Mhash/W (2x5870).

There are a lot of assumptions here, but given the trend of processing power increase this is something to consider before jumping into mining (unless you already have the hardware).

Of course the big risk here is that BTC price can move wildly. I think a $20/BTC is a reasonable estimate for the average price in the next few months - but far from certain of course. However, we MUST recognize that there is no basis for difficulty of mining to have any effect on the price of a coin in the short term.

Difficulty adjusts according to the processing power of the network, which keeps the rate of coin generation roughly at 2000 per 2 weeks. So regardless of the amount of processing going on out there, the supply of new BTC is fixed per time unit. Since the total supply of BTC is essentially fixed, the dollar price is only influenced by the demand for BTC. There is no way to know what the demand side is going to do, and what kind of political risk we are facing (U.S. govt could outlaw it for example).

As a result, this calculation is inexact at best. But I think I've presented at least a plausible scenario given the data. Let me know if you differ in opinion.

You can't verify that a specific person sent the money to you by matching addresses. That's just the way the system works. See this discussion: http://forum.bitcoin.org/?topic=2273.0

The workaround to this problem is that the recipient sets up a new receive address for every client. That way, when money shows up in the account, the recipient knows who it's from.

I wanted to reply to a thread in the miners section...but instead had to make one here in the noobies section. Can I please get out of here!?

I promise I'm not a bot or threadcrapper.

Best settings for 5850 on win7, SDK 2.4 are:

phoenix miner, -k phatk VECTORS BFI_INT AGGRESSION=13 FASTLOOP=false WORKSIZE=256

it nets me 340mhash with 850 core. About 10mhash more than any other setup I've tried.

Give it a try.